Reproductive risk factors of breast cancer in elderly population. A single centre study

Abstract

21187 Background: Relationship between reproductive variables and cancer risk has been yet established. Protective effect of late age at menarche (≥ 11) and high parity (≥ 4) and the increased risk of nulliparous women, older age at first birth (AFB) (≥ 28) and older age at menopause (≥ 52) seems to act differently in pre vs. postmenopausal elderly women. The aim of this study was to describe recognized reproductive risk factors for breast cancer in a selected population of a single centre. Methods: A cross- sectional study of women diagnosed ≥ 70 years. We selected women with removed early breast cancer and on follow-up. No disease recurrences were allowed. Between January 2005 to June 2006. Described reproductive risk factors for breast cancer were collected. Tumour characteristics also. Results: 91 patients were recruited. Mean age at time of diagnosis: 76 (70–92); mean age at study 80 (71- 95). Reproductive factors: late age at menarche (12; 9–16; percentile 75: 14) and late age at menopause: 48 (29–56); 75% ≥ 52 years old were common. Nulliparous women were similar at described in other series (10%); parity 2 (0–9); percentile 75: 3;late AFB: 26 (19–40); percentile 75: 29 years. No one had taken hormonal treatment. Only 16% had familiar history of breast cancer and 9% personal history. Almost all (84%) had positive tumours for the oestrogen receptor and 76% for progesterone receptor. 91% had ductal histology and 54% had II stage. Conclusions: Discussion: unexpected associations between classical reproductive risk factors and breast cancer on elderly patients have been described in other series. It seems that late age at menarche and late age at menopause is common in elderly breast cancer population, as in our serie. Nulliparous elderly continue to have risk but seems that the protective effect of parity dissipated in these population. Older age at first birth (≥28) is consistently associated with increased breast cancer risk. Conclusion: our study is based on a high selected population and clearly has a selection bias. Our results cannot be generalized to other populations but describes our patients. The results therefore should be internally valid. And it seems to coincide with other studies. No significant financial relationships to disclose.